Re: 19991018: Bulletins seperators

On Mon, 18 Oct 1999, Unidata Support wrote:

> >To: address@hidden
> >From: address@hidden
 > >Subject: Bulletins seperators
> >Organization: .
> >Keywords: 199910181909.NAA15792
> 
> TO: address@hidden
> 
> 
> Hello,
> 
> I am currently connecting to an LDM as a server.  I.e.
>  - connected as a client
>  - request what I want to receive via a Hyia and feedme request.
>  - reverse my connection to server.
> 
> I have looked into the LDM code on how to seperate standard WMO bulletins   
> but
> every once in a while I receive "badly" form of bulletins separator.
> 
> In the pqing source code the standard bulletins encapsulators are:
> 
> SOH CR CR LF   for a bulletin start
> CR CR LF ETX   for a bullentin end.
> 
> Now I receive from time to time "seperators" like
>
Claude,

Are you saying the LDM is rejecting these products with messages about
badly formed seperators?  The seperators have been changing on NOAAport
HDS for the past 8 months. It seems that the changes are occuring when the
model programs have been updated or they are moved to  run on new
types of hardware. The new seperators are still within the WMO
conventions so they are legal. I expect that the seperators will continue
to change for the next couple of years. No, I haven't found any place
that states the changes or why the seperators have changed. The decoders
have to be flexible enough to accept the new seperators. If you download
the Unidata decoders package and look at code for gribtonc, the code
demostrates how the start and end of a GRIB product can be determined 

Robb...




 
> SOH CR LF or 0xFE CR LF for a bulletin start
> CR LF ETX   for a bulletin end
> 
> Is there somewhere a place that defines all (or most) of the bulletins   
> seperators
> that I can rely on...
> 
> 
> I also would like to separate binary data (radars, ...) from the same   
> feed.  GRIB
> bulletins are "encapsulated" in standard WMO seperators (SOH, CR CR LF   
> ...
> CR CR LF ETX) but other binary data doesn't seem to be.  Is there an easy   
> way
> to know where a binary chunk of data starts and end from a feed comming   
> out of
> an LDM and if there is a sequence of characters that marks the beginning   
> and the
> end of it.
> 
> Thank you in advance for your help
> 
> Claude Lamy
> email: address@hidden
> 

===============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
address@hidden             WWW: http://www.unidata.ucar.edu/
===============================================================================